Pancreatic ductal adenocarcinoma (PDAC) is the cause of death of over 300,000 people per year worldwide. PDAC has a very poor prognosis; only 5% of diagnosed PDAC patients survive more than five years. The carcinoma is commonly diagnosed during late stages, and surgery to resect the tumor is only possible in about 1/5th of new cases. There exists an urgent need to develop new therapeutic drugs for PDAC, yet pancreatic research progress has been slowed by a lack of understanding of the causes of PDAC. In particular, the role of pancreatic stroma in tumor development and promotion has been of intense interest to cancer biology research and pharmacological research. The pancreatic stroma consists of a diverse and heterogeneous population of cancer associated fibroblast (CAF), immune, vasculature, stellate, mast, and nerve cells along with their associated extracellular matrix and soluble proteins. Several mechanisms for stromal involvement in PDAC have been hypothesized, but to study stromal involvement in PDAC, disease models for this cancer must be made to include the stroma in a biologically relevant manner.Recent technologies have laid the foundation for combining stromal cells within a PDAC tumor model. It has been shown that primary human PDAC tissue samples can be cultured to form pancreatic organoids. Pancreatic organoids have distinct advantages over other pancreatic cancer tumor models such as patient- derived xenografts (PDXs) or genetically engineered mouse models (GEMMS) and have a great potential as in vitro disease models. Microraft array cell sorting technologies have emerged as effective tools for manipulating populations of cells on a single cell level. This research proposes to utilize microraft arrays to sort mixtures of primary human cancer-associated fibroblast stromal cells and tumor cells to generate pancreatic organoids with organized, controlled fibroblast milieus. To accomplish this, a high-throughput automated pancreatic cell culture, imaging, and sorting system based on microraft arrays will be developed. It will first distinguish between tumor cells, normal CAFs, and activated CAFs in culture on microrafts using a novel immunostaining assay for known biomarkers. Then, it will sort a defined proportion of tumor, CAF, and activated pro-tumor CAF cells into each well of an organoid culture array plate. Each organoid site cultured in the plate will thus contain a controlled normal/activated CAF microenvironment in interaction with a tumor cell population.The innovative automated cell sorting system proposed in this research will be an enabling technology providing an unparalleled control over the composition of pancreatic organoid microenvironments. Using this system, distinct components can be chosen from the incredibly diverse stromal population and cultured in the microenvironment of developing tumor organoids. This research can vastly increase the potential to use pancreatic organoids as PDAC disease models to guide cancer research and therapy development.